> if you think all miles of flight are the same

OP said air safety is going down broadly [1]. So yes, considering all air transport miles is valid given the context.

[1] https://news.ycombinator.com/item?id=38929237

more accidents is less safe, anyway you cut it. I really can't understand how you can view flying as not being less safe in this moment. If you want to do integration over huge time spans to make your point, lets start at zero and go to infinity... human lifespan is 0 years long average over history, seeing as we didn't exist for some period of time. So any changes to human life in shorter time spans is completely meaningless to an average. Is this 0 year lifespan a useful statistic?

> how you can view flying as not being less safe in this moment

There has been one passenger fatality aboard a major U.S. airline since November 2001.

Actually, if we average flight fatalities starting at the year zero, the average fatalities for all flying planes is zero!

> more accidents is less safe, anyway you cut it. I really can't understand how you can view flying as not being less safe in this moment

You’re proposing crashes are autocorrelated. They’re not. They would be in a vacuum filled with spherical cows. But grounding and investigating takes care of that.

This is related to the fallacy of thinking if a coin has come up heads thrice in a row, it’s more likely to come up tails the fourth time. It’s not. We have a lot of innocuous flight miles as data from which to make robust statements, particularly when it comes to characterising the safety of the entire airline industry.

Put another way, on which day are you safer, the day before the accident or the day after?